Sample · Client details anonymized · Structure and output are unchanged
FORMA
Session Recap
Good to meet you both.
Here is the recap we promised, landing about half an hour after we hung up. Treat it as a bit of a meta moment. It was curated by AI and then fine tuned by a human, which is the working relationship we spend most of our time teaching.
This is the agentic AI you brought up as one of your gaps. What you are reading right now was built by an agent. The lift on our end was small, and it is the kind of thing your team could be doing inside a week.
Everything we covered is below.
01
Roughly 340 people, mostly in Chicago, with about 210 in the office and 130 remote. The office week splits Tuesday and Wednesday, then Thursday and Friday, which is why the pilot below is built as two sessions rather than one.
You are running LLM agnostic on an internal platform and leadership has already made AI fluency a priority. Between that mandate and the two of you in people operations, the conditions here are better than most places we walk into.
02
Skill Gap
One of the two things named as the biggest gap. Prompts are already being written inside Notion, and the AI feature there was just switched on.
Skill Gap
The other big one. Your internal AI team has built some, but the wider staff has close to no training on it.
Past Training
You have taken other courses and spoken to other companies. Every time, it stayed abstract and never tied back to the work you actually do.
Format
The async sessions did not land. Lecture rooms of 60 people meant nobody asked a question, which is where the real learning was supposed to happen.
Blocker
Leadership gets nervous about note takers and AI tools sitting in on calls. This came up as the part that keeps stalling things.
03
Brett and Yo came out of the same industry you work in, so the exercises get built on your work instead of generic examples. We have worked both the agency side and the vendor side, and that shared ground is most of why this has landed across media and tech.
No more than ten minutes goes by without people working in the tool themselves. Laptops open, real outputs, everyone watching what changes when they change the input.
We sit with your stakeholders before any session, walk through their day, find the bottleneck, and design the exercises around what we hear. We do not come in blind and we do not assume.
Prompt engineering, contextual engineering, and hallucination mitigation. About 85 percent of it is model agnostic, which fits the internal platform you already run on.
Everyone leaves with a third party accredited certificate in AI Fluency and Foundational Prompt Engineering.
The engagement does not stop at the last session. Open Q and A on real problems people are stuck on, the same way we run it with our enterprise clients.
04
Decision 01
A full day immersive at the Chicago office for the people who come in, with a virtual option so the 130 remote staff are not cut out. You both said in person is where your people actually pay attention, and we agree.
Decision 02
Wednesday for one cohort and Thursday for the other, matching how your office already splits the week. Roughly 20 people per room.
Decision 03
Rather than a company wide mandate, start with a group small enough to see the value immediately and let it travel by word of mouth. Senior managers came up as the likely first cohort.
Decision 04
Since there is no budget set yet, we will come to you with our guidance on what it takes to get a minimum viable group trained properly. A proposal follows by Friday with a recommended number and scope, so you have something concrete to react to and take upstairs.
05
FORMA · By Friday, July 17
Two full days in person at your Chicago office, Wednesday and Thursday groups, about 20 each. It will include our recommended investment for a minimum viable pilot and a recommended number of discovery hours, so there is something on paper for you to react to.
FORMA · This week
Dates for the next public session, which lands early in the fall. You both want the course under your belt before recommending it internally, and that is the right instinct.
FORMA · Once the pilot is approved
We will need a named stakeholder or two from whichever cohort goes first, so we can build the exercises around their real bottlenecks.
Your team · Early fall
Take the open workshop yourselves so you are recommending something you have been through rather than something you read about.
Your team · Timing up to you
Once the proposal is in hand, make the internal recommendation. Tell us if it helps to have Brett or Yo on that conversation and we will join.
06
Four things we left unresolved. None of them hold up the proposal, but your answers will sharpen it.
The proposal comes to you by Friday. React to it, push back on it, cut it in half if you want. It exists to give you something concrete to work from.
Anything we missed, reply here and we will pick it up. Thanks for the time today.
Yo Shahar and Brett Whelan
FORMA Media